It's an interactive map showing live generation for major wind farms in Great Britain, showing what each wind farm is generating both now and in the past, and where that generation is physically located.
Animated weather data is from WeatherLayers and shows current and historic wind conditions on the map, providing context to wind generation around the country.
History mode allows you to go back in time and see wind generation and weather conditions for a particular date and time. It's great for exploring days of record generation, such as the 21.6GW record on January 10th, 2023!
Prediction mode lets you see what wind farms are estimated to be generating using current wind conditions and model based on historic generation and wind speeds. Is a wind farm generating as you expect, or is there something to look into?
An experimental feature allows you to see what future wind farms could be generating today (or in the past!) if they were already built and operational. If you click the "sparkle" button on the map you'll get to see what the upcoming 3.6GW Dogger Bank wind farm is estimated to generate if it was operational today.
There's an "About" section on the site that goes into detail on the various public data sources and how some of the features work. I also document a lot of this on my Twitter @robhawkes if you're curious.
This is just the start and there are many more features to come!
Please let me know your comments and suggestions.
Do you also have access to solar data? Eg, I am a UK homeowner w/ a recent solar installation on my roof, would be curious to see how that looks too.
The interface isn't great, but i find it pretty useful sometimes.
There's some previous work in this area by others so I imagine my near future will be spent digging through various papers on the subject.
I found sheffieldsolar.co.uk but could not find any data within that site.
A ~4kW installation gives me up to ~25kWh on a good day in summer. The big factor is actually the precession of the seasons since I'm at 56 degrees north. Retail electricity prices are now over 30p a kWh, so anything you can do to offset that is very useful.
We do get the occasional sunny day contrary to popular belief, but even on cloudy days it's possible to generate a decent amount of energy.
We also have very long days in summer which can make for a decent amount of kWh/day (although the opposite is true in winter)
So it's not so sunny, especially as you go north, but films/TV do exaggerate the problem. "I'm from London" regularly gets a comment about terrible weather, "I'm from Paris" does not.
There's very little solar power in Scotland, as you'd expect: https://electricityproduction.uk/plant/solar/map/
(Figures from https://en.wikipedia.org/wiki/London#Climate and so on.)
Then there's the likes of Ripple Energy's solar park co-op offering. The levelised cost of buying in is around 2.5p kWh, with another 2p kWh going on maintenance. The return on the 1st (and active) co-op agreement there via a wind farm is 28p kWh, no reason why the solar park would give different returns than the wind one. Here in the UK grid pricing is more or less the same nationally.
Unfortunately there's much less of a return in Winter months due to higher latitudes, though in Summertime the days are longer than in many other countries.
There's a great site [1] that calculates total solar generation around the UK using a mix of real data and forecasts. We were generating at 5GW around midday and that number will only increase.
It's been fascinating to watch the various weather patterns over time to see where the gaps are in the locations of wind farms around the UK. It'll be great once we fill in the big hole in capacity off the south-west coast.
My casual observation is that it’s generally windy somewhere around the coast, and when not in one area like the north sea it is windy of the south west or south coast. And vice versa.
Once we build out farms around the four axis of the coast im sure the base load generated in aggregate will rise quite a bit.
https://en.wikipedia.org/wiki/Wind_power#/media/File%3AGloba...
The UK has had issues with limited transmission line capacity to Scotland which had led me to think that most of the generation potential was up there.
The Scotland thing is interesting as there's definitely issues with transmission constraints over the B6 boundary, though there's still benefits to building more wind farms for local consumption and/or storage as more batteries (especially) and pumped hydro capacity is added to the mix. Plus the National Grid have plans to expand capacity through to 2030 but I had to admit that's at the limit of my knowledge.
Every time I look at anything to do with grid scale energy, you see just how important it is to distribute and diversify everything from sources of energy to geographic location to interconnects.
Capitalism with its capacity utilisation targets, just in time delivery etc is not very well suited for a robust network.
Is that a problem with the data accuracy, or is that reflective of the real-world situation (e.g. misalignment of the turbines against the wind)?
(p.s. thanks for your work at Octopus!)
- Wind speeds are from GFS forecasts and cover quite a wide area - The realtime generation data from the balancing system isn't 100% precise and is averages over a 30-minute period - Some wind farms are simply much more efficient, especially large offshore ones - Individual turbines might be down for maintenance and/or not running efficiently - Localised issues like wake effects and turbulence
I'm not an expert in all this though so take it with a pinch of salt – I'm just very interested in it all and love making visualisations about it.
All that said, one of the intents behind creating the map is to highlight exactly this sort of stuff and to spark a discussion around why it might be.
Therefore in high wind, you would expect that in areas of dense generation, especially those with limited capacity interconnects to the rest of the grid, curtailment may be seen.
(nb of course you might overprovision it to some degree on the presumption that, say, only 70% of your generation is usable at any one time; I more mean the numbers like say 16%.)
WeatherLayers consists of two products, that can be used either together or separately:
- WeatherLayers GL is a frontend visualisation library with deck.gl layers which can be integrated with common map libraries. The library can be used either with custom self-hosted data or with WeatherLayers Cloud.
- WeatherLayers Cloud is a cloud service providing pre-processed data for visualization from common public weather data sources (NOAA, Copernicus).
Athough it’s a commercial project, there are ways to discuss a potential discount or even a free usage for a non-commercial project such as Robin’s.
Finally, a cloud service that has met it's true calling :)
I have a random question - what layer are you using to draw the country borders on the base map? It really stands out, and helps tie the styling together.
The countries layer is using this dataset from MapTiler [1] which is also the service I'm using for all the base maps and vector tile data. I then used interleaved layers [2] in Deck.gl / MapLibre to place the wind layers behind the country outlines.
[1] https://www.maptiler.com/countries/ [2] https://deck.gl/gallery/mapbox-layer
deck.gl focuses on vector data, whereas WeatherLayers GL adds raster capabilities. Actually, WeatherLayers GL is mostly custom WebGL shaders that could be implemented outside of deck.gl as well. The major reason why deck.gl is used is that it serves as an integration layer for popular map libraries. I've written deck.gl-leaflet [1] plugin myself.
The real eye-opener (for me) on the latter display is the emissions per kwh over all time - dropped steadily from 500g in (only!) 2012 to a steady 150g or so recently. Encouraging!
I'm going to look at how to be a little more accurate with this, similar to Kate's great website.
However, their main drawback is that they are standalone, can’t be integrated to another application (besides an iframe).
WeatherLayers is built as a tool for developers, to allow for a seamless integration to other applications and allow overlaying with custom data such as wind power plants in this case, to support end-users with understanding the data in natural context.
https://app.electricitymap.org/map?wind=false&solar=false
https://github.com/electricitymaps/electricitymaps-contrib/b...
Beautiful project you've done, really enjoy the visualizations.
(no association, just a fan)
This Serverless Function has timed out.
Your connection is working correctly.
Vercel is working correctly.
504: GATEWAY_TIMEOUT
Code: FUNCTION_INVOCATION_TIMEOUTHoping it'll be resolved soon.
I'd wonder if you have a lot of trees if you'd ever get much output from a turbine anyway. At a certain point, if you want legitimate solar or wind power, you probably end up finding a spot to clear some tree cover. Grab a pallet of dirt cheap used panels from SanTan and throw 'em out on the ground.
Personally I probably wouldn't even think about wind unless I lived out in eastern Oregon or Washington on the plains, with big enough property that I could get a halfway decent size turbine and mount it far enough from the house so I couldn't hear it.
My idea was to utilize a combination of solar and this.
For example, PVWatts [1] makes it very easy to get solar potential for a site (at hourly granularity). But having trouble finding something similar for wind. Ideally want something that is also validated against data from real world wind farms after they're built.
With that said, some of the wind turbines already reach 190m in height (eg. Hornsea One) and the mega-sized Haliade-X turbines at the upcoming Dogger Bank wind farm are 260m tall!
Grid have access to "power available" signals, but this is not public data.
This is partly why I've been collaborating with someone on a prediction feature that tries to guess generation using realtime wind conditions and historic actual generation data at various wind speeds.
I'm hoping to improve things as time goes on and I learn more about the underlying data (and its quirks).
I'm hoping to expand beyond GB and Ireland would be a great next step if the data is publicly available.
The only comment I have is that concentric circles are not the most intuitive way of showing a percentage of a whole. 50% doesn't 'look' like 50%, and looks similar to 30%.
but I'm not sure how else you would do it.